364 ANNALS NEW YORE ACADEMY OF SCIENCES 



duction of the dermal rays has frequently occurred in the Actinopterygii 

 (e. g., Murcena, Lepidosiren) without impairing the locomotive power of 

 the fish. When the pectoral fin was turned downward and forward the 

 medial surface of the spreading lobate portion of the fin would seem to 

 be fitted for the digital emargination of the borders and the modification 

 of the radials to form the digits. Dr. H. H. Wilder (1909, p. 235) has 

 also called attention to the fact that in the manus of Necturus the 

 muscles for spreading and closing the digits are so highly developed as 

 to suggest derivation from a primitive appendage which had the power 

 of widely spreading and closing, as in the fins of fish. In Sauripterus 

 (Fig. 10) and Eusthenopteron (Fig. 8) most of the radials converge 

 toward the mesopterygial axis, which includes the supposed homologues 

 of the humerus, ulnar, ulnare and digit Y ( ?), and the same is true of 

 the digits of the manus of Carboniferous Amphibia (Fig. 11) and Eep- 

 tilia and of recent urodeles; in both classes the post-axial paramere, or 

 radius, is sharply separated from the remainder of the cheiropterygium. 

 I find that both Emery (1897, p. 208) and Jaekel (1909) have also 

 noted this biramous character of the tetrapod cheiropterygium. 



The muscles of the pectoral limbs of Sauripterus were probably sep- 

 arable into a deep proximal mass covering the scapulocoracoid and the 

 humerus and a deep distal mass running from the humerus to the distal 

 radials. Surface muscles perhaps extended from the radials back to the 

 scapulocoracoid region. The joint corresponding to the elbow joint is 

 clearly present in these rhizodonts and the "radius'^ and "ulna" are seen 

 to be a part of the outer or distal segment of the limb. 



The pelvic limb of Eusthenopteron (figured by Goodrich, 1909, p. 275), 

 which is the best known one of the rhizodonts, has likewise a certain re- 

 semblance to the tetrapod t3^pe, as noted by Jaekel (1909), in so far as 

 it possesses a single basal piece analogous to the femur and two main 

 radials analogous to the tibia and fibula. The pelvic limb, like the 

 pectoral, appears to represent an imperfectly attained mesorhachic type 

 in which the preaxial elements have become more or less produced later- 

 ally and to some extent regrouped. The "femur" has two broad distal 

 facets for the supposed fibula and tibia. The mesopterygial element, 

 which may be called the "fibula," is a wide element bearing at its distal 

 end two facets for the proximal "tarsals" which appear to be the fibulare 

 and intermedium, or tibiale. The distal end of the supposed tibia, or 

 preaxial paramere, lies beside the supposed intermedium; all the tarsals 

 slant toward the fibular or mesopterygial axis. The joints corresponding 

 to the knee and ankle joints are well defined (Fig. 11). All this con- 

 stitutes a distinct resemblance to the pelvic limbs of Carboniferous Tetra- 



